$ terminal

The CLI Power Guide

34 AI skills. One command. Every workflow from idea to production.

cm — Cody the Hamster 🐹

    ( . \ --- / . )
     /   ^   ^   \
    (      u      )
     |  \ ___ /  |
      '--w---w--'

Cody v4.1.3

34 Skills · Ship 10x faster

Dashboard: ● RUNNING http://localhost:6969

? Quick menu: › ↑↓ navigate · Enter select · Ctrl+C exit

❯ 📊 Dashboard — Open in browser

📋 My Tasks

📈 Status

🧩 Browse Skills

➕ Add a Task

⚡ Install/Update Skills

❓ Help

// CHEAT SHEET

All 34 Skills at a Glance

Type /cm- in your AI chat to access any skill. Copy any command below and paste it directly.

Command	Category	What It Does
/cm-start	Command	Autonomous mode — AI picks skills based on your goal
/cm-status	Command	View progress summary of current tasks
/cm-dashboard	Command	Open visual Kanban board to track work
/cm-planning	Engineering	Explore intent, write implementation plan before coding
/cm-tdd	Engineering	Write tests first, then implement (RED → GREEN → REFACTOR)
/cm-debugging	Engineering	5-phase root cause analysis before fixing
/cm-execution	Engineering	Batch, subagent, or parallel execution modes
/cm-code-review	Engineering	Review lifecycle — request, handle feedback, merge
/cm-git-worktrees	Engineering	Isolated git worktrees for feature branches
/cm-continuity	Operations	Working memory across sessions via CONTINUITY.md
/cm-terminal	Operations	Enforce progress logging and error-stop behavior
/cm-skill-chain	Operations	Compose skills into automated pipelines
/cm-skill-mastery	Operations	When to invoke skills, how to create new ones
/cm-deep-search	Operations	Semantic codebase search for large projects
/cm-identity-guard	Security	Verify correct account before push/deploy
/cm-secret-shield	Security	Pre-commit secret scanning, blocks leaked keys
/cm-safe-deploy	Security	8-gate deploy pipeline with rollback
/cm-quality-gate	Security	Evidence-based verification before completion
/cm-test-gate	Security	Set up reliable test infrastructure
/cm-safe-i18n	Security	Safe multi-language translation with 8 audit gates
/cm-brainstorm-idea	Product	Multi-dimensional analysis with 9 Windows + Double Diamond
/cm-ux-master	Product	Design system with 48 UX Laws, Harvester v4
/cm-ui-preview	Product	Visual UI preview via Google Stitch / Pencil
/cm-jtbd	Product	Jobs-To-Be-Done customer discovery
/cm-project-bootstrap	Product	Full project scaffold from day 0
/cm-dockit	Product	Auto-generate docs, SOPs, API references
/cm-content-factory	Growth	AI content engine with multi-agent support
/cm-ads-tracker	Growth	Full pixel + CAPI tracking setup
/cro-methodology	Growth	Conversion audit + A/B test design
/cm-readit	Growth	Audio reader + voice CRO triggers
/cm-how-it-work	Command	Full workflow guide from idea to production
/cm-skill-index	Command	Progressive skill discovery (3-layer loading)

// SCENARIOS

Common Workflow Recipes

Copy-paste these into your AI chat. Each recipe chains skills in the right order.

🚀 New Project (Full Lifecycle)

new-project.sh

# Verify identity first
$ /cm-identity-guard

# Scaffold everything
$ /cm-project-bootstrap

# Plan your first feature
$ /cm-planning "User auth with email + OAuth"

# Build with TDD
$ /cm-tdd

# Deploy when ready
$ /cm-safe-deploy

🐛 Bug Fix (Root Cause First)

bugfix.sh

# Isolate in a worktree
$ /cm-git-worktrees fix/login-crash

# Find the real cause
$ /cm-debugging "Login page crashes on submit"

# Write regression test, then fix
$ /cm-tdd

# Review and merge
$ /cm-code-review

🌍 i18n Translation Batch

i18n.sh

# Translate a page to all languages
$ /cm-safe-i18n "Translate pricing.html to vi, zh, ko, hi, ru"

# Outputs:
✓ Extracted 47 strings
✓ Translated to 5 languages (parallel)
✓ Passed 8/8 audit gates
✓ Updated en.json, vi.json, zh.json...

🎨 UI Redesign

redesign.sh

# Analyze current design
$ /cm-brainstorm-idea "Improve conversion on pricing"

# Design with UX laws
$ /cm-ux-master

# Preview before coding
$ /cm-ui-preview

# Plan the implementation
$ /cm-planning

// FIRST RUN

What Happens When You Run `cm`

First-time users get a 5-step guided setup. Cody walks you through name, platform, first task, skill discovery — coding in under 2 minutes.

cm — First Run Experience

    ( . \ --- / . )
     /   ^   ^   \
    (      u      )
     |  \ ___ /  |
      '--w---w--'

Hi there! I'm Cody 🐹

Your smart coding companion.

Step 1 of 5 ● ○ ○ ○ ○

◇ What should I call you?

│ Tody

✓ Nice to meet you, Tody!

Step 2 of 5 ● ● ○ ○ ○

◆ Where do you code?

│ ● ✦ Google Antigravity (Gemini) recommended

│ ○ 🟣 Claude Code · ⬡ Cursor · 🌊 Windsurf · 🔶 Cline

│ ○ 📦 OpenCode · 🪁 Kiro · 🤖 Copilot · 🔧 Other

✓ Great choice! gemini

Step 3 of 5 ● ● ● ○ ○

◇ What are you working on?

│ Landing page

✅ First task created! 🤖 "Landing page"

Step 4 of 5 ● ● ● ● ○ — 34 superpowers discovered! 🧩

Step 5 of 5 ● ● ● ● ● — 🎉 You're all set!

── Quick Reference ──

cm Quick menu · cm status Progress · cm list Skills · cm dashboard Mission Control

// BEST PRACTICES

Power User Tips & Exceptions

Level up your workflow and learn how to handle edge cases.

⚡

Combine Skills in One Prompt

Reference multiple skills: @cm-brainstorm-idea @cm-ux-master @cm-planning — AI chains them automatically in the right order.

🔄

Resume After a Break

Start every new session with /cm-continuity. AI reads CONTINUITY.md and picks up exactly where you left off — no context lost.

🌿

Isolate Risky Work

Use /cm-git-worktrees before any large refactor. Creates a separate directory so your main branch stays clean.

🏃

Parallel Execution Exception

If a task fails during full autonomous mode, switch to /cm-execution in parallel mode to isolate subagents to single tasks.

🛡️

Pre-Push Safety

Always run /cm-identity-guard + /cm-secret-shield before pushing. This catches wrong-account deploys and leaked API keys.

❌

Context Overload Loop

If AI starts looping or forgetting rules, STOP. Do not keep prompting. Run /cm-continuity to summarize, then start a fresh chat session.